Joseph Cabinta

Airbnb Revenue Analysis

SQL    Power BI

Digging into Airbnb’s revenue and operations data using SQL and Power BI to spot trends, catch anomalies, and offer some suggestions.

Despite a rise in bookings and listings, 2022 saw a dip in revenue. This analysis looks into why, highlights high-performing spots like Big Bear Lake and Yuca Valley, and suggests ways to bounce back through smarter pricing.

Apple Products Performance Analysis

SQL    Power BI

In this analysis, I explored Apple’s product sales from 2019 to 2022 to understand how shifts in pricing, product mix, and geographic demand shaped performance. Using SQL and Power BI, I tracked the rise of high-end models, the short-lived surge of accessories like Airtags, and the strong performance of U.S. cities like Brooklyn and New York. The data reveals how Apple’s strategy leaned toward premiumization—even as quantity sold declined.

Amazon Sentiment Analysis

Python    Power BI

A recent study by Sameh Al-Natour and Ozgur Turetken, published in the International Journal of Information Management, found that textual sentiment provides stronger predictive power for future sales than star ratings alone. They highlight how reviews carry richer emotional and contextual signals, offering a deeper lens into customer satisfaction.

Inspired by this, I analyzed Amazon’s Jan 2023 product data to explore sentiment trends across categories and reveal what the stars might be hiding. The project uncovered hidden gems with high sentiment but low visibility and flagged bestsellers with surprisingly poor sentiment trends.

Aerosol Formation Descriptors

Python    SQL

Tiny particles known as aerosols greatly influence our climate and health by shaping weather patterns, regulating temperatures, and impacting air quality. Despite this, we still don’t fully understand the exact steps or energy changes involved in how these particles form and grow from gas molecules. This study focuses on two key processes: nucleation, where gas molecules stick together to form tiny clusters, and growth, where these clusters expand by attracting more molecules.

Using molecular simulations, we explored how different gases interact at the microscopic level and how cluster shapes and compositions evolve. Our findings provide new insights to improve models for predicting air quality and climate changes.

Materials for Water Purification

Python    SQL

Cadmium in water is dangerous to both people and the environment, so I studied how to remove it using special sponge-like materials called MOFs. I used computer simulations to test different versions of a material called UiO-66, focusing on how changes in its structure affect cadmium removal. I found that adding more carboxylic acid groups made the material much better at capturing cadmium, with one version removing up to 88% while absorbing less water. This shows that we can design better water filters by adding the right chemical groups that attract toxic metals like cadmium.